Research Journal of Applied Sciences, Engineering and Technology 4(22): 4748-4754, 2012
ISSN: 2040-7467
© Maxwell Scientific Organization, 2012
Submitted: April 03, 2012
Accepted: April 23, 2012
Published: November 15, 2012
Performance Comparison of IEEE 802.11e EDCA and 802.11b DCF Under NonSaturation Condition using Network Simulator
1
1
G. Prakash and 2P. Thangaraj
Department of ECE, K.S.Rangasamy College of Technology, Namakkal, Tamilnadu, India
2
Department of CSE, Bannariamman Institute of Technology, Erode, Tamilnadu, India
Abstract: In this study, throughput and delay performance of IEEE 802.11b and 802.11e is presented under
non-saturation conditions. In order to improve the performance of IEEE 802.11b, the IEEE 802.11e has been
proposed to improve the Quality of Services (QoS) for multimedia application. The standard 802.11b
CSMA/CA contention mechanism does not support QoS but the standard 802.11e provides QoS by adjustment
of MAC parameters. The comparison of 802.11b DCF and 802.11e EDCA mechanism by using Network
Simulator (NS-2) with different parameters such as throughput, delay, CWmin and AIFS differentiation are
simulated. The EDCA stations have more competitive advantages than 802.11b under all the above parameters.
The simulation results show that the proposed algorithm improves the performance of the EDCA stations.
Keywords: DCF and EDCA, IEEE 802.11, non-saturation, WLAN
INTRODUCTION
IEEE 802.11 Wireless Local Area Network (WLAN)
(IEEE Standard 802.11, 1999) is one of the most widely
deployed wireless network technologies in the world
today. Distributed Coordination Function (DCF) and
Enhanced Distributed Channel Access (EDCA) are the
fundamental access mechanisms for IEEE 802.1b and
IEEE 802.11e (IEEE Standard 802.11, 1999; IEEE
Standard 802.11, 2005) respectively. The DCF and EDCA
implement a Binary Exponential Backoff (BEB)
algorithm by increasing the contention window size
exponentially for each transmission failure for collision is
resolved (IEEE Standard 802.11, 1999; IEEE Standard
802.11, 2005; Bianchi, 2000). Therefore, the focus of this
study is to modify BEB for improving the performance of
DCF and EDCA stations using network simulator (NS 2
version 2.29). Many researchers have done individually
by study of IEEE 802.11 DCF and EDCA performance in
both analytically and simulation. Most of them assumed
as an ideal channel condition, which means that the
packet corruptions are only due to collision (Bianchi,
2000; Vassis and Kormentzas, 2005; Kong et al., 2004).
But few of them assumed that non ideal channel condition
which means that packet collision due to noise
(Dhanasekaran and Krishnan, 2010; Daneshgaran et al.,
2008). Previous researches of our study (Prakash and
Thangaraj, 2011a; Prakash and Thangaraj, 2011b; Prakash
and Thangaraj, 2010) have analyzed non-saturation
throughput performance of the IEEE 802.11 DCF and
EDCA in the presence of transmission error, but few of
them done comparison of DCF and EDCA under non
saturation traffic condition.
In real network, traffic is mostly non-saturation
(mobile stations have not always packet to transmit). In
this study, we extend the previous studies for the
comparison of DCF and EDCA with different parameters
such as number of stations, CWmin differentiation,
throughput, media access delay and AIFS differentiation.
We invite the interested reader to refer the basics of DCF
and EDCA of IEEE 802.11 is presented in Prakash and
Thangaraj, (2011b) Prakash and Thangaraj, (2010)
Prakash and Thangaraj (2011c) and Kong et al., (2004).
The simulator considers an Infrastructure BSS (Basic
Service Set) with an AP and a certain number of mobile
stations which communicates only with the AP. For
simplicity, we assume that data packets transmitted by
different stations are involved by the same probability of
error. This way, channel errors on the transmitted packets
can be accounted for as it is done within ns-2 (network
simulator-ns-2, 2010). The simulation results show that
the proposed scheme provides a remarkable performance
improvement in WLAN environments.
MATERIALS AND METHODS
The detailed differences between DCF and EDCA: The
DCF is designed for best-effort data transmission by using
Carrier Sense Multiple Access with Collision Avoidance
(CSMA/CA). The DCF scheme does not provide any
Corresponding Author: G. Prakash, Department of ECE, K.S. Rangasamy College of Technology, Namakkal, Tamilnadu, India
4748
Res. J. Appl. Sci. Eng. Technol., 4(22): 4748-4754, 2012
Fig. 1: Proposed binary exponential backoff workflow for IEEE 802.11b DCF
means of service differentiation and thus assumes that all
flows have equal priority. The main concern of DCF is to
reduce the collision among the flows that are competing
for access to the wireless medium. In DCF, the backoff
counter is decremented at the end of each slot following
DIFS. On the other hand, the backoff counter is
decremented after AIFS has passed in EDCA scheme
(Hwang and Cho, 2006; Majkowski and Palacio, 2006;
Bianchi et al., 2005). In DCF, a transmission can begin if
backoff counter makes a transition to 0 and the medium is
idle. Differently, EDCA station can transmit only if a
backoff value is already 0. In EDCA, the backoff counter
can be decremented or the station starts Rx toTx
turnaround when backoff counter value is zero.
While DCF and PCF schemes were not able to fulfill
the QoS requirements for multimedia applications. DCF
is simple and allocate wireless medium access to all flows
in the same manner. PCF, though it includes service
differentiation mechanisms, it still considers all flows
from a specific station to have the same priority.
4749
Res. J. Appl. Sci. Eng. Technol., 4(22): 4748-4754, 2012
Therefore, the 802.11e WLAN standard has introduced
the Enhanced Distributed Channel Access (EDCA), which
adds transmission prioritization to CSMA/CA. EDCA is
a completely distributed scheme and allows each station
to sort its traffic in four different Access Categories (AC)
(Kong et al., 2004; Engelstad and Osterbo, 2005; Xiao,
2005; Qiang, 2005; Vassis and Kormentzas, 2005). By
doing this, EDCA provides service differentiation, taking
into consideration the various needs of flows within a
specific station. As such, EDCA could be considered as
the new version of the legacy DCF.
Modified backoff work flow: In this study, we modified
the binary exponential backoff algorithm for the case of
transmission or channel error. In the noisy wireless
environment, without distinguish packet collision and
channel error, the DCF cannot adjust the backoff
procedure properly. To resolve this problem, we present
new backoff algorithm. When a collision occurs, the
backoff time of the collided stations are doubled to reduce
the contention. When there is a channel error due to
channel noise, instead of doubling the contention window
as in the standard, the station select the backoff time from
the same contention window. Simulations results show
that the new backoff algorithm working significantly
improve the throughput in WLAN. Therefore the backoff
algorithm will cause long delay and poor channel
utilization when there are successive transmission errors.
This backoff algorithm is explained in the flow chart
shown in Fig. 1.
Another effective modification of the EDCA has been
proposed in the event of transmission errors. In the basic
AC, the contention window is doubled after every
unsuccessful transmission. Unsuccessful transmission
happens in two cases:
C
C
Internal collision or virtual collision of a packet with
other packets with in a station
Due to error in the channel or packet collision with
other station
Here the internal collision refers to the collision
among the queues of different priorities inside a station,
while the external collision refers to that among different
stations. We proposed Binary Exponential Backoff
workflow for IEEE 802.11e EDCA for considering
internal collision and erroneous transmission under nonsaturation traffic conditions, instead of doubling the
contention window in the internal collision the backoff
counter selects a counter value from the same contention
window. This backoff algorithm is explained in the flow
chart shown in Fig. 2. We consider a WLAN in
Unsaturation (Nonsaturation) condition, that is buffer of
the transmitting station is empty, after a successful
transmission.
CWmin and AIFS differentiation: EDCA is designed to
provide prioritized QoS by enhancing the contentionbased DCF. Before entering the MAC layer, each data
packet received from the higher layer is assigned a
specific user priority value. EDCA introduces four
different First-In First-Out (FIFO) queues, called Access
Categories (ACs). Each data packet from the higher layer
along with a specific user priority value should be mapped
into a corresponding AC according to a Table 1 (Hwang
and Cho, 2006; Majkowski and Palacio, 2006;
Bianchi et al., 2005). Different kinds of applications (e.g.,
background traffic, best effort traffic, video traffic and
voice traffic) can be directed into different ACs. Each AC
behaves as a single DCF contending entity with its own
contention parameters (CWmin[AC], CWmax[AC], AIFS[AC]
and TXOPLimit[AC]), which are announced by the QAP
periodically in beacon frames.
The CWmin differentiation employed in EDCA is to
change the amount of TXOPs provided to each traffic
class. A station with a lower value of CW will reduce the
average time needed to successfully deliver a packet and
thus experience improved performance in comparison to
stations with higher CW values. The average value of the
CW can be tuned through differentiated setting of the
backoff parameters and specifically of CWmin and CWmax.
The CWmin differentiation employed in EDCA is to
change the amount of TXOPs provided to each traffic
class. A station with a lower value of CW will reduce the
average time needed to successfully deliver a packet and
thus experience improved performance in comparison to
stations with higher CW values. The average value of the
CW can be tuned through differentiated setting of the
backoff parameters and specifically of CWmin and CWmax.
AIFS differentiation is to reserve channel slots for the
access of higher-priority stations. This is accomplished by
using different AIFS values for different traffic classes.
The AIFS is the amount of time a station defers access to
the channel following a busy channel period. Once an
AIFS has elapsed, the station access is managed by the
normal backoff rules (Hwang and Cho, 2006; Majkowski
and Palacio, 2006; Bianchi et al., 2005).
A basic issue of AIFS differentiation is that confined
slots occur after every busy channel period. This implies
that the percentage of confined slots significantly
increases as long as network congestion increases. In
reality, a greater number of competing stations involves
that the average number of slots between consecutive
busy channel periods reduces and thus the fraction of
protected slots over the total number of idle slots gets
larger.
4750
Res. J. Appl. Sci. Eng. Technol., 4(22): 4748-4754, 2012
Fig. 2: Proposed binary exponential backoff workflow for IEEE 802.11e EDCA
Table 1: Default parameters for EDCA
Voice
Video
Transport protoco lUDP
UDP
AC
VO
VI
CWmin
3
7
CWmax
7
15
AIFSN
2
2
Packet size
160 bytes 1280 bytes
Packet interval
20 ms
10 ms
Sending rate
64 kb/s
1024 kb/s
SIMULATION RESULTS OF DCF AND EDCA
Background (best effort)
UDP
BE
15
1023
3
1500 bytes
12.5 ms
960 kb/s
Wireless Local Area Network (WLAN) consists of
two different set of STAs, one set of STAs is running
under IEEE 802.11b MAC protocol and another set of
STAs is running under IEEE 802.11e MAC protocol.
EDCA stations have been configured with the standard
DCF backoff parameters (CWmin = 31 and CWmax = 1023).
The packet size has been fixed to 1024 bytes and the
4751
Res. J. Appl. Sci. Eng. Technol., 4(22): 4748-4754, 2012
CWmax
aCWmax
aCWmax
aCWmin
aCWmin/2
0.7
AIFSN
7
3
2
2
EDCA
0.6
0.5
normalised throughput
Table 2: EDCA default settings
Access category
CWmin
AC_BK
aCWmin
AC_BE
aCWmin
AC_VI
aCWmin/2
AC_VO
aCWmin/4
0.7
EDCA
normalised throughput
0.6
0.4
0.3
DCF
0.5
0.2
0.4
0.1
DCF
0.3
0
EDCA with AIFSN=1
EDCA with AIFSN=2
EDCA with AIFSN=3
0
5
10
15
number of stations
20
25
0.2
Fig. 4: DCF vs. EDCA throughput with AIFS differentiation
0.1
0
0
EDCA with CWmin=7
EDCA with CWmin=15
5
10
15
number of stations
20
0.8
0.7
25
normalised throughput
0.6
Fig. 3: DCF vs. EDCA throughput with CWmin differentiation
retransmission limit is set to 7 for all the stations. Control
frames are transmitted at a basic rate equal to 1 Mbps,
while the MAC Protocol Data Unit is transmitted at 2
Mbps. Table 2 shows the default values of the channel
access parameters defined in EDCA for the four ACs (BK
= background, BE = best effort, VI = video, VO = voice).
In order to be granted priority over the DCF stations,
EDCA must be configured with CWmin values smaller than
the legacy DCF value CWmin = 31. The Network Simulator
NS-2.29 version (network simulator-ns-2, 2008) is used
for simulation.
Figure 3 shows that for low values of number of
stations (n), the involvement between EDCA and DCF
stations is contrariwise related to the employed CWmin
value. For example, in the case of n = 10, the throughput
performance of EDCA when CWmin = 7 is about double
times the corresponding throughput performance of DCF
(which uses CWmin = 31); similarly, when CWmin = 15, it
is about double the DCF throughput. In general, the
throughput performance of EDCA is proves better
performance result as shown in the Fig. 3. As the number
of competing stations grows, the EDCA throughput
significantly reduces, while the DCF also decreases.
Figure 4 shows that comparison of DCF and EDCA
throughput with AIFS differentiation simulation results for
AIFSN = 1, 2, 3 and DCF stations coexist as the number
of station for each scheme increases. As the number of
stations for each scheme increases, the difference of
throughput also increases because more EDCA stations
get chances to transmit a packet due to one decrement of
backoff counter at the end of AIFS after every
0.5
0.4
0.3
0.2
0.1
0
802.11e
802.11b
0
2
4
6
8
10
12
Stations
14
16
18
20
Fig. 5: Throughput comparisons for DCF vs. EDCA stations
transmission. The total throughput decreases as the
number of stations increases, which is a consequence of
increased collisions
Figure 5 presents the throughput result and
reconfirms that 802.11e provides remarkably improved
throughput in comparison to 802.11. In Fig. 5 shows that
EDCA has enhanced throughput when compare to DCF,
because the 802.11e stations have a high priority to
transmits a packet than DCF stations.
Figure 6 shows that throughput with different ACs
for EDCA and DCF stations. We observe that throughput
of EDCA stations for voice and video traffic is
abnormally improved as compared to 802.11b.The
throughput increases for higher priority (voice and video)
and low throughput for low priority (best effort and
background). It shows that throughput for all four traffic
streams starts to drop equally as the sixth and eighth
stations are added to the network in 802.11. On the other
hand, 802.11e provides traffic prioritization (voice and
video) through its service differentiation mechanism. The
4752
0.8
0.8
0.7
0.7
0.6
0.6
Media acc ess delay (sec )
normalised throughput (Mbps)
Res. J. Appl. Sci. Eng. Technol., 4(22): 4748-4754, 2012
0.5
0.4
0.3
voice 802.11e
video 802.11e
Best effort 802.11e
Background 802.11e
voice 802.11b
video 802.11b
Best effort 802.11b
Background 802.11b
0.2
0.1
0
0
2
4
6
8
10
S tations
12
14
Fig. 6: Throughput comparisons with
categories for DCF and EDCA
16
0.5
0.4
0.3
AC2
0.2
AC1
DCF
AC0
0.1
18
20
0
different access
5
10
15
20
25
30
Number of Stations
35
40
45
Fig. 8: Media access delay versus the number of stations in
non-saturation
12
AC3
media acc ess delay (ms )
11
and compared with DCF. We observe that the medium
access delay for the middle priority stations (AC-2) is
very close values of DCF case. This means that the
enhancement in the delay for high priority stations.
Furthermore, the MAC delay increase for all ACs after 10
(pkts/sec). But delays in high priority (AC 0 and AC1)
have significant improvement.
Figure 8 shows the non-saturation delay versus the
number of stations for different access categories with
DCF. High priority classes have a small media access
delay that has a significant impact in the low priority
classes. Middle priority classes are not affected. As it is
shown, the increase in the traffic of AC-2 stations affects
mostlyAC-0 stations. AC-1 stations have a similar
performance to that of the DCF case and AC-2 stations
obviously perform better than in the DCF case.
AC2
10
DCF
AC1
9
AC0
8
7
6
5
4
3
2
5
10
15
packet generation rate (pkts/s)
20
Fig. 7: Media access delay versus packet generation rate
throughput decreases for best effort and background
traffic streams in 802.11e compared to 802.11b stations.
Finally, the comparison shows that 802.11e offers
improved service to higher priority traffic than low
priority traffic which is poor performance than 802.11b.
In addition the throughput versus stations for different
access categories comparison with DCF stations. We
observe that EDCA is effective in providing service
differentiation in terms of throughput, higher priority
AC’s always perform better than lower priority ones.
In addition to higher priority AC’s gets saturates
after, AC’s 0 and 1 (voice and video) saturates for n>6
with higher throughput, while AC’s 2 and 3 (Best effort
and background) saturate for n>8. From this figure the
throughput of DCF is significantly increases with AC’s 2
and 3 priorities station.
Figure 7 show the media access delay versus the
packet generation rate for a random station of each ACs
CONCLUSION
This research present study on performance of
802.11b DCF and 802.11e EDCA. The contention-based
EDCA mechanism can provide effective service
differentiation between different types of traffic. The
performance is measured by non-saturation traffic
conditions. The simulation output is shown that EDCA
scheme has much better performance over DCF stations
especially at low traffic load. Since EDCA stations get
more chances to transmit a packet due to one decrement
of backoff counter at the end of AIFS after every
transmission. The total throughput decreases as the
number of stations increases, which is a consequence of
increased collisions. The Network Simulator (NS-2.29)
tool is used for simulation. Analysis of jitter and delaycontrol will be investigated for the future study.
4753
Res. J. Appl. Sci. Eng. Technol., 4(22): 4748-4754, 2012
REFERENCES
Bianchi, G., I. Tinnirello and L. Scalia, 2005.
Understanding 802.11e contention-based
prioritization mechanisms and their coexistence with
legacy 802.11 stations. IEEE Network, 19: 28-34.
Bianchi, G., 2000. Performance analysis of the IEEE
802.11 distributed coordination function. IEEE J.
Select. Areas Commun., 1: 535-547
Dhanasekaran, S. and A. Krishnan, 2010. Non-saturation
throughput enhancement of IEEE 802.11b distributed
coordination function for heterogeneous traffic under
noisy environment. Int. J. Autom. Comput., 7(1):
95-104.
Daneshgaran, F., M. Laddomada, F. Mesiti and M.
Mondin, 2008. Unsaturated throughput analysis of
IEEE 802.11 in presence of non-ideal transmission
channel and capture effects. IEEE T. Wireless
Commun., 7(4): 1276-1286.
Engelstad, P. and O. Osterbo, 2005. Delay and throughput
analysis of IEEE 802.11e EDCA with starvation
prediction. Proceedings of the 30th IEEE Conference
on Local Computer Networks, pp: 647-655
Hwang, G.H. and D.H. Cho, 2006. Performance analysis
on coexistence of EDCA and legacy DCF stations in
IEEE 802.11Wireless LANs. IEEE T. Wireless
Commun., 5(12): 3355-3359.
IEEE Standard 802.11, 1999. Wireless LAN Medium
Access Control (MAC) and Physical Layer (PHY)
Specifications.
IEEE Standard 802.11, 2005. Wireless LAN Medium
Access Control (MAC) and Physical Layer (PHY)
Specifications: Medium Access Control (MAC)
Quality of Service (QoS) Enhancements.
Kong, Z.N., D.H.K. Tsang, B. Bensaou and D. Gao,
2004. Performance analysis of IEEE 802.11e
contention-based channel access. IEEE J. Select.
Areas Commun., 22(10): 2095-2106.
Majkowski, J. and F. Palacio, 2006. Coexistence of IEEE
802.11B and IEEE 802.11E stations in QoS enabled
wireless local area network. Proceedings of the
Communication Systems and Applications (CSA)
Conference. pp: 102-106.
Prakash, G. and P. Thangaraj, 2010. Throughput analysis
of IEEE 802.11b WLAN under a Non-Saturated
condition. Proceedings of the International
Conference on Recent Trends in Information,
Telecommunication and Computing IEEE Computer
Society, pp: 65-69.
Prakash, G. and P. Thangaraj, 2011a. Non-saturation
throughput analysis of IEEE 802.11 distributed
coordination function. Europ. J. Scient. Res., 51(2):
157-167.
Prakash, G. and P. Thangaraj, 2011b. Analytical
modeling of IEEE 802.11e enhanced distributed
channel access under a non-saturation condition.
7(4): 554-560.
Prakash, G. and P. Thangaraj, 2011c. Throughput analysis
of IEEE 802.11e EDCA under Non Saturation
Condition. Proceedings of the International
Conference on Electronics Computer Technology
IEEE Explore. pp: 117-121.
Qiang, N., 2005. Performance Analysis and
Enhancements for IEEE 802.11e Wireless Networks
IEEE Network: The Network Simulator-ns-2.
Retrieved From: http:/www.isi.edu/nsnam/ns/,
(Accessed on: June 26, 2010).
Vassis, D. and G. Kormentzas, 2005. Delay performance
analysis and evaluation of IEEE 802.11e EDCA in
finite load conditions. Wireless Pers. Commun., 34:
29-43.
Xiao, Y., 2005. Performance analysis of priority schemes
for IEEE 802.11 and IEEE 802.11e wireless LANs.
IEEE T. Wireless Commun., 4(4): 1506-1515.
4754